POSTGRAD. MED. J. (1965), 41, 425

CAPILLARY PERMEABILITY TO PLASMA PROTEINSA. W. ASSCHER, M.D., B.Sc. (Lond.), M.R.C.P.(Lond.).J. HENRY JONES, M.D. (Wales), M.R.C.P. (Lond.).

Medical Unit, Welsh National School of Medicine, Cardiff Royal Infirmary.

ACCORDING to the classical hypothesis ofStarling (1896) the exchange of water andsmall lipid-insoluble molecules across the capil-lary membrane is largely governed by thebalance between hydrostatic and colloid osmoticpressures, the existence of the latter beingdependent upon relative impermeability of thecapillary membrane to plasma proteins. Star-ling's concept has been verified by directexperimental measurements (Landis, 1927;Pappenheimer and Soto-Rivera, 1948) and itsessential features are no longer in dispute. Inrecent years, however, attention has been drawnto the fact that considerable quantities ofplasma proteins pass through the capillarywalls. Drinker and Field 1(1931) collected lymphfrom the cervical, leg and renal lymphatics, aswell as from the thoracic duct, and foundconsiderable concentrations of protein in allsamples; they concluded that "the capillariespractically universally leak protein" and thatthis protein can only return to the bloodstream via the lymphatic system. This conceptof a continuous exchange of plasma proteinsbetween plasma and extravascular compart-ments has been confirmed repeatedly withradioactively labelled plasma proteins (e.g.,Wasserman and Mayerson, 1951; Yoffey andCourtice, 1956). 13I-labelled albumin turnoverstudies in the human subject indicate an over-all exchange rate between intra- and extra-vascular pools of 117-234% (mean 140%) ofthe total plasma albumin per day; and thetotal amount of albumin in the extravascularpool is greater than that in plasma (Cohen,Freeman and McFarlane, 1961). Leakage ofplasma proteins may be of importance in thatit enables antibodies to reach the tissues, andcertain hormones such as insulin to exert theirdirect effect on the tissues (Hechter and Lester,1960). Other hormones (e.g., thyroxine andcortisol) are partly bound by plasma proteins,and many drugs are known to be bound byplasma albumin. There is also evidence thatthe presence of plasma proteins in the inter-stitial fluid plays a part in the maintenance of

normal tissue electrolyte content (Carey andConway, 1954; Creese and Northover, 1961)and that the large extravascular reserve ofplasma protein is drawn upon when protein israpidly removed from the blood stream (Was-serman, Joseph and Mayerson, 1956; Cope andLitwin, 1962).

In this article the mechanism of proteinleakage through capillary walls, the methodsavailable for measuring the leakage, and thefactors that control it will be reviewed.

Theoretical ConsiderationsThe permeability of a membrane is difficult

to define. Landis (1946) considered that, atthe very least it should be defined in termsof volume or mass of substance passingthrough unit area, in unit time, under theinfluence of unit hydrostatic or unit osmoticpressure. In practice, however, it is usuallynot possible to measure all these variables, andwhen considering the permeability of a mem-brane to large molecules, like proteins, theratio of the concentration of the particularsubstance on either side of the membrane (thefiltrate/filtrand ratio) is an acceptablesimplification.

All the available evidence suggests thatplasma proteins, and other molecules of similarsize, pass through the capillary membrane atrates inversely related to the size of the mole-cule (see Yoffey and Courtice, 1956; Landisand Pappenheimer, 1963). The similarity bet-ween capillary filtrates and those obtained bythe use of inert artificial membranes led to the"Pore Theory" of capillary permeability. Thispostulates that the capillary membrane con-tains numerous minute holes, or pores, whichallow rapid passage of small ions and lipid-insoluble molecules but restrict the rate ofmovement of larger molecules such as proteins.Landis (1946) suggested that the capillary wall

consists of a meshwork with pores of manydifferent sizes, the diameter of the largestpores being at least 38 A. A 'graded poretheory' such as this would satisfactorily account

by copyright. on D

ecember 18, 2020 by guest. P

rotectedhttp://pm

j.bmj.com

/P

ostgrad Med J: first published as 10.1136/pgm

j.41.477.425 on 1 July 1965. Dow

nloaded from

POSTGRADUATE MEDICAL JOURNAL

for 'the inverse relationship between rate ofpassage and molecular size. Pappenheimer(1953), however, in a detailed critical analysis ofthe pore theory of capillary permeability, cameto the conclusion that the facts could best beexplained on -the basis of "restricted diffusion"through an isoporous membrane; it was cal-culated that uniform cylindrical pores witha diameter of 60-90 A and a population densityof 1-2 x 109/cm2 would account for the observedrates of passage of water and lipid-insolublemolecules of various sizes.

Large molecules, like proteins, may traversea porous membrane in one of two ways. Theymay be carried either by bulk-flow in a streamof other molecules (filtration) or by diffusiondue to movement of the protein moleculesamongst the other molecules. Both bulk-flowand diffusion contribute to the leakage of pro-teins from the vascular compartment, but therelative importance of these two mechanismsis uncertain, and may differ in different areasand at different rates of filtration. Accordingto Pappenheimer (1953) the passage of proteinmolecules, whether by diffusion or bulk-flow,is restricted by at least two forces, namelyviscous drag (or friction) between the moleculesand the walls of the pores, and steric hindranceat the en'rance to a pore (steric hindranceimplies that the molecule can only enter apore if it does not strike the edges of the pore).The magnitude of these forces is determinedby the relationship between the size of theprotein molecule and the size of the pore;the larger the molecule, the greater the resist-ance to its passage. Viscous drag and sterichindrance, therefore, determine the selectivityof protein leakage and account for "molecularsieving" through an isoporous membrane; thecomplex physical factors involved are discussedat length by Pappenheimer (1953) and Landisand Pappenheimer (1963).

It has been pointed out by Courtice andhis colleagues (Courtice and Morris, 1955;Yoffey and Courtice, 1956; Courtice, 1961) thatpores of the diameter suggested by Pappen-heimer (1953) would not explain the escape ofthe large ,/-lipoprotein molecule into inter-stitial fluid. Similarly, the detailed studies ofGrotte (1956) using dextran fractions of varioussizes indicated the presence of 'pores' of atleast 120 A radius. Grotte postulates two differ-ent pore systems viz. capillary pores ofrelatively uniform size (40-60 A radius) andcapillary 'leaks' (radius 120-130 A); an approx-imate ratio of capillary 'leaks' to capillary'pores' was calculated to be 1:34,000. (Grotte,

1956; Areskog, Artusan, Grotte and Wallenius,1964). Results similar to those obtained byGrotte (1956) have been reported by Mayerson,Wolfram, Shirley and Wasserman (1960). Theseauthors, also, concluded that the data indicatetwo discrete sets of pores-a set of smallpores allowing the passage of dextran mole-cules not greater than 250,000 M.W., andlarge pores permitting the passage of mole-cules of at least 412,000 M.W. An alternativeexplanation suggested is that molecules of mole-cular weight greater than 250,000 are trans-ferred by a process called 'cytopempsls' (seebelow) rather than by passage through pores.Both Grotte (1956) and Mayerson and others

(1960) observed regional differences in capillarypermeability, the hepatic capillaries being part-icularly permeable. It was suggested that largepores (or 'leaks') predominate in the liver,and that regional differences are due to differ-ences in the relative proportion of small poresand 'leaks'.

Pappenheimer (1953) considered the poresof the capillary membrane to be rigid struc-tures. This concept may need to be modified,since the observations of Shirley, Wolfram,Wasserman and Mayerson (1957) suggest thatcapillary distension produced by intravenousinfusion of albumin greatly increases the poro-sity of the membrane. These authors attributedthis phenomenon to stretching of the pores,and concluded that the pores were not rigidstructures.

The Anatomical Basis of Capillary PermeabilityDuring the last decade, the electron micro-

scope has revealed striking regional variationsin the structure of capillaries (see Bennett,Luft and Hampton, 1959). The following sec-tion is a brief account of the electron micro-scopic structure of capillaries and summarisesthe attempts which have been made to correlatestructure with function.The capillary endothelium. The endothelium

of capillaries may be fenestrated or non-fenestrated. Non-fenestrated endothelium is en-countered in the capillaries of muscle, myo-cardium, lung and nervous system. In thesecapillaries, adjacent endothelial cells areseparated by very narrow tortuous channelswhich measure some 100 A in width. Thecytoplasm of the endothelial cells adjoiningthese channels is condensed to form 'attachmentbelts' or 'desmosomes'. Fenestrated endotheliaare of two types-the fenestrae may be intra-cellular or intercellular. Intracellular fenestrae,measuring 300-600 A in diameter, are found

426 July, 1965by copyright.

on Decem

ber 18, 2020 by guest. Protected

http://pmj.bm

j.com/

Postgrad M

ed J: first published as 10.1136/pgmj.41.477.425 on 1 July 1965. D

ownloaded from

July, 1965 ASSCHER and JONES: Capillary Permeability to Plasma Proteins

in the endothelium of the glomerular andtubular capillaries of the kidney, in the in-testine and in several endocrine glands. Rhodin(1962) showed that these fenestrae are nottrue pores but that they are closed by thinmembranes which probably represent attenua-tions of the endothelial cells. Luft and Hechter(1957) showed that fresh ox adrenal-glandcapillaries did not have intracellular fenestra-tions whereas if fixation was delayed for twohours, fenestrations were apparent. Moreover,it was found that if adrenal glands which hadbeen left for two hours were perfused withwarm ox blood, the fenestrations could bemade to disappear. It is possible, therefore,that intracellular fenestrations only occur as aresult of anoxia. Intercellular fenestrations upto several thousand Angstrom units in dia-meter occur in the endothelium of liver andspleen capillaries. These are real pores andtheir presence may well explain the high pro-tein content of hepatic lymph (Yoffey andCourtice, 1956).

Palade (1953 and 1956) and Moore andRuska (1957) demonstrated that the endothelialcells of non-fenestrated endothelia containednumerous pinocytotic vesicles which measuredsome 200-600 A in diameter. These authorsalso noted invaginations of the cell membraneboth on the luminal and outer aspect of theendothelial cells (caveolae). It was suggestedby Palade (1953) that this system of vesiclesand caveolae might act as a transport mechan-ism, the caveolae being nipped off to formvesicles and the vesicles discharging their con-tents on the outer surface of the endothelialcell after they had migrated across the endo-thelial cell cytoplasm. This method of transportwas referred to as pinocytosis by Palade and,more correctly, as transport by cytopempsis byMoore and Ruska. The existence of such atransport system is supported by the observa-tions of Wissig (1958) who showed that ferritinmolecules injected intravenously were incorpo-rated in pinocytotic vesicles and finally appearedin the pericapillary space and basement mem-brane. Palade (1960) made similar observationsusing intravenous colloidal gold. Further sup-port for transport by cytopempsis was obtainedby Alksne (1959) who showed that under theinfluence of histamine, a substance known toincrease vascular permeability to plasma pro-teins (vide infra), the incorporation of colloidalgold and mercuric sulphide particles intopinocytotic vesicles was much increased. Florey(1964) who studied the perfused heart showedthat colloidal ferric hydroxide particles entered

the caveolae and became incorporated invesicles. It seems likely that transport bycytopempsis would require energy and it issurprising, therefore, that Florey (1964) wastotally unable to influence the incorporationof colloidal ferric hydroxide into pinocytoticvesicles by the addition of respiratory inhibitorsto the fluid perfusing the isolated heart or bygradually cooling the perfusion fluid. Nor didthese procedures alter the number of pino-cytotic vesicles. He concluded, therefore, thatit seems unlikely that the vesicles are beingactively formed and destroyed and that theyare probably more permanent structures thanhad previously been thought. There is still nodirect proof that transport by cytopempsisoccurs.The basement membrane. The basement

membrane forms a continuous coat ofmoderate electron density some 200-500 Athick. It does not represent condensed 'cement'substance, for when a capillary is damagedand the endothelial cells retract from it, thebasement membrane remains behind as adistinct layer. Palade (1953) regarded thebasement membrane as a felt of fine fibrilswhose meshes are filled with an amorphousmatrix. It has been claimed that the basementmembrane may act as a filter. There is noexperimental proof for this, except perhaps inthe case of glomerular basement membranes.Farquhar, Wissig and Palade (1961) showedthat the main filtration barrier to ferritin mole-cules in the rat kidney was the basementmembrane and that in rats with amino-nucleoside nephrosis ferritin molecules pene-trated the basement membrane more speedily.The adventitial layer. This layer forms the

outer coat of capillaries and consists of fusi-form cells (pericytes) together with collagenand reticulin fibres. It may be complete, asin the capillaries of the brain and spinal cord,or incomplete, as in the capillaries of cardiacand skeletal muscle, liver and kidney (Bennettand others, 1959). There is no evidence thatthe adventitial layer contributes in any way tothe 'semi-permeability' of the capillary mem-brane.Two conclusions can be drawn from the

foregoing electron microscopic observations:1. Although functionally the capillary mem-

brane behaves as a porous membrane, electronmicroscopy has failed to show the existenceof pores except in the liver and spleen. It maybe, of course that in the living capillary suchpores do exist but until techniques are devisedfor obtaining the requisite high magnification

427by copyright.

on Decem

ber 18, 2020 by guest. Protected

http://pmj.bm

j.com/

Postgrad M

ed J: first published as 10.1136/pgmj.41.477.425 on 1 July 1965. D

ownloaded from

POSTGRADUATE MEDICAL JOURNAL

for studying living capillaries this questioncannot be settled.

2. Cytopempsis as a mechanism of trans-port across the capillary wall remains unproven.It is possible, however, as suggested by Mayer-son and his colleagues (1960), that macromole-cules are transported in this way.Methods of Assessing Capillary Permeability

It is extremely difficult, if not impossible,to obtain a true sample of normal capillaryfiltrate, and for this reason the methods avail-able for measuring capillary permeability toplasma proteins are indirect.An early technique that has been used ex-

tensively measures the local accumulation inthe tissues of proteins labelled with a dye orwith 131iodine. Plasma albumin binds vital dyessuch as Evans' blue, trypan blue or pontamineblue, and in animals with recently injected dyein their blood an intradermal injection of asubstance that increases capillary permeabilityresults in local 'blueing' of the skin due toincreased leakage of dye-stained protein(Rocha e Silva and Dragstedt, 1941; Menkin,1940; Miles and Miles, 1952). Dye containing3ZBr may also be used (Cope and Moore, 1944)and vascular permeability in normal andoedematous brain tissue has been investigatedwith fluorescein-labelled proteins (Klatzo,Miquel, and Otenasek, 1963). These methods,however, are not quantitative, and may bemisleading. Measurement of the amount ofdye in tissues is made difficult by the needto remove blood from the tissue without alsoremoving tissue fluid (Cope and Moore, 1944),while the 'skin-blueing' method requires thatthe injection of the test substance does notalter blood flow in the local vascular bed(Miles and Miles, 1952).

Landis, Jonas, Angevine and Erb (1932)compared blood samples taken simultaneouslyfrom the brachial veins of congested and con-trol arms and showed that the permeabilityof the capillary wall to protein varies with thedegree of venous congestion. Under standardconditions it may be possible to gain someinformation about capillary permeability bythis method but, as Landis and his colleaguesemphasized, "caution must be used in extend-ing conclusions based on observations involvinghigh grades of venous congestion to considera-tions of normal capillary permeability".

Drinker and 'his colleagues believed that theprotein concentration of lymph and of tissuefluid in any area is approximately the same(Drinker and Field, 1931; Drinker, Field, Heim

and Leigh, 1934; Drinker and Yoffey, 1914).It is now recognized, however, that the proteinconcentration in the lymph draining from anarea does not necessarily reflect the permeabilityof the capillaries in that area, since the proteinconcentration in interstitial fluid, and thereforein lymph, might increase as a result of increasedabsorption of water at the venous end of thecapillary without any change in capillary per-meability. Conversely, a decreased amount ofprotein might be found in the lymphatic trunkin the presence of increased capillary per-meability if the blood flow were sufficientlyreduced (Cope and Moore, 1944; Yoffey andCourtice, (1956). Measurements of the totalamount of protein drained from a given areamight be considered a better index of per-meability but alterations in haemodynamics,and particularly in the number of capillariesbeing perfused at the time make this, too, anunreliable method.Whi'e the measurement of total protein con-

centration in interstitial fluid or lymph is ofno value in determining capillary permeabilityuseful information may be obtained by measur-ing the relative concentrations of proteins ofdifferent sizes. Thus, an increase in the size ofthe pores reduces molecular-sieving, and leadsto a relative increase in the amount of largeproteins in the filtrate. It was early recognizedthat not only albumin but also globulins andfibrinogen pass through the capillary wall, andthat for most regions capillary permeability toalbumin was greater than that to globulins(Drinker and Yoffev. 1941). Perlmann, Glennand Kaufmann (1943) examined lymph pro-teins by Tiselius electrophoresis and foundthe same four electrophoretic components asare present in serum. The albumin/globulinratio in lymph exceeded that of plasma and,moreover, it was shown that the distributionof proteins in lymph changed when capillarypermeability increased following burs; theinjury permitted both an increased passage ofplasma proteins and a decrease in the differ-ential permeability to albumin. Courtice andMorris (1955) and Courtice (1961), usingpaper electrophoresis, confirmed that thealbumin/globulin ratio of lymph was greaterthan that of plasma, and demonstrated thateven the largest protein, -namely /8-lipoprotein(Molecular weight 1,300,000), escapes into thetissue fluid.Hardwicke and Squire (1955), investigating

glomerular permeability, performed simultane-ous quantitative analysis of urine and serum inpatients with proteinuria and expressed the

July, 1965428by copyright.

on Decem

ber 18, 2020 by guest. Protected

http://pmj.bm

j.com/

Postgrad M

ed J: first published as 10.1136/pgmj.41.477.425 on 1 July 1965. D

ownloaded from

ASSCHER and JONES: Capillary Permeability to Plasma Proteins

results in terms of the separate clearances ofalbumin, a,, a2, / and y globulins. The clear-ances of the four globulin fractions were thenexpressed as a percentage of the albuminclearance, and in this way it was possible toclassify proteinuria according to its selectivity.It was argued that if tubular reabsorption ofprotein is non-selective (and evidence to sup-port this was produced) then the degree ofselectivity of the proteinuria reflects changes inglomerular permeability. The available evidencesuggests that the lymphatic membrane readilyallows the passage of all plasma proteins and,indeed, even larger particles such as red cells(for review see Yoffey and Courtice, 1956 andMayerson, 1963). It seems likely, therefore, thatprotein reabsorption into lymphatics is non-selective with regard to molecular size, andthat measurement of differential protein 'clear-ance' into tissue fluid on lymph might providea good index of capillary permeability.* Thetechnique of paper electrophoresis is not idealfor this purpose, however, since the electro-phoretic components, apart from albumin, arenot homogeneous with regard to molecular size(,-globulins, for example, include siderophilinof molecular weight 90,000 and /-lipoprotein,of molecular weight 1,300,000). Immunologicalmethods have allowed more precise measure-ments of the renal clearance of a number ofindividual proteins in patients with proteinuria(Soothill, 1962; Squire, Hardwicke and Soothill,1962); similar methods may be used in analys-ing tissue fluids.The development of gel-filtration techniques,using 'Sephadex' (Porath and Flodin, 1959and Flodin, 1962) has provided a relativelysimple method of separating proteins accordingto their molecular size. Hardwicke (1965) hasused this technique to measure differentialprotein clearances into urine. The method mayalso be used for 'measuring differential clear-ances of plasma proteins into oedema fluid andascites,(Jones and Peters, 1965).

After an intravenous injection of 13I-labelledalbumin, plasma radio-activity falls rapidly atfirst, because labelled albumin is being removedby diffusion into the extravascular space aswell as by catabolism. Later, when mixing oflabelled albumin in the plasma and extravas-cular pools is complete, the rate of fall inplasma radioactivity is reduced, since diffusionfrom plasma is now balanced by the return oflabelled albumin from extravascular sites viathe lymphatics. The shape of the plasmaradioactivity curve may, therefore, provide*Capillary porosity is, strictly, a more accurate term.

some indication of the rapidity of mixing, andof the relative quantities of albumin in theintravascular and extravascular compartments.A mathematical analysis widely used for thesepurposes is that described by Matthews (1957).Results obtained by this method of analysisagree well with those derived by other tech-niques (Cohen and others, 1961), but there isno proof that intercompartment transfer ratesderived in this way are an indication of capillarypermeability, and the dynamics of plasma/extravascular exchange are analysed in termsof the total amount of protein present in eachcompartment rather than in terms of proteinconcentration.Much useful information has resulted from

the use of polydisperse macromolecules otherthan plasma protein,-in particular from studieswith dextrans. Glomerular permeability has beeninvestigated by measuring the clearance ofdextran fractions of different mean molecularsize. Brewer (1951), Wallenius (1954) and Was-serman, Loeb and Mayerson (1955) determinedplasma/lymph concentration ratios for dextransof different molecular weight in an attemptto measure capillary permeability. Later, (May-erson and others, 1960) the method wasimproved by infusing dextran fractions and'11I-labelled serum albumin simultaneously."Steady state" lymph/plasma concentrationratios for each dextran fraction could then berelated to the ratio for labelled albumin anda coefficient obtained which referred the per-meability of a capillary bed with respect to agiven molecular weight of dextran to its per-meability with respect to albumin. Grotte andhis colleagues have also made extensivestudies of capillary permeability using dextranfractions (Grotte, 1956; Areskog and others,1964). Dextran fractions of sufficiently uniformmolecular size for studies of this nature aredifficult to prepare, and the methods used bythese authors for estimating the molecularsize and concentrations of dextrans in the bio-logical fluids are tedious. Recently, Brooks,Davies, Graber and Ricketts (1960) havesucceeded in labelling dextran with radio-iodide,thus enabling small quantities to be measuredaccurately; while gel-filtration techniques,using 'Sephadex', allow the distribution ofmolecular sizes present to be determined. Theserecent developments eliminate the need toprepare dextran fractions of 'narrow' mole-cular weight distribution and enable moreprecise work to be done with dextran (Hard-wicke, Hulme, Jones and Ricketts, 1964).

Grotte, Juhlin and Sandberg (1960) have used

July, 1965 429

by copyright. on D

ecember 18, 2020 by guest. P

rotectedhttp://pm

j.bmj.com

/P

ostgrad Med J: first published as 10.1136/pgm

j.41.477.425 on 1 July 1965. Dow

nloaded from

430 POSTGRADUATE MEDICAL JOURNAL July, 1965TABLE 1

SUBSTANCES WHICH INCREASE VASCULAR PERMEABILITY TO PLASMA PROTEINSAminesHistamine Lewis (1927), Gaddum (1948), Spector and Willoughby

(1957), Majno and Palade (1961).5-Hydroxytryptamine Rowley and Benditt (1956), Majno and Palade (1961).

Peptides'Leukotaxine' Menkin (1956), Duthie and Chain (1939), Cullumbine

and Rydon (1946), Spector (1951).Bradykinin Holdstock, Mathias and Schachter (1957), Rocha e Silva

(1955).Kallidin Hilton and Lewis (1955), Werle (1955).Substance P. Euler and Gaddum (1931), Perow (1955).Substance U. Beraldo (1955).PepsitensinPepsitocin Croxatto (1955).PepsanurinKinin Holdstock, Mathias and Schachter (1957).Angiotensin Giese (1964).

ProteinsPermeability Factor (Dilution) Wilhelm, Miles and Mackay (1955), Mill, Elder, Miles

and Wilhelm (1958), Stewart and Bliss (1957).Permeability Factor (Age). Miles and Wilhelm (1955).Permeability Factor (Native) Elder and Wilhelm (1958).Permeability Factor formed by Mill (1964).interaction of guinea-pig tissuewith serum.Lymph Node Permeability Factor Walter and Willoughby (1965).Kallikrein Frey, Kraut and Werle (1950), Werle (1955).Renin Winteritz, Mylon, Waters and Katzenstein (1940).Masson, Plahl, Corcoran and Page (1953), Asscher and

Anson (1963 a and b), Giese (1963).

solid spherical particles (radius 300-700 A) ofmethylmethacrylate, marked with a fluorescentdye, to determine the maximum particle-sizethat can cross the 'blood-lymph barrier' atdifferent sites in the body.

It is evident that there is no satisfactorymethod of measuring capillary permeability innormal subjects. Apparently satisfactory tech-niques of collecting normal extracellular fluidhave been described in animals (Creese,D'Silva and Shaw, 1962) but not in man. Inoedematous subjects, however, samples ofextracellular fluid are easier to get. Quantitativeanalysis of the proteins present in such fluidby the more precise methods now availablewill probably be useful.

The Regulation of Capillary Permeability withRespect to Plasma ProteinsThe factors which account for changes in

capillary permeability have been studiedmainly in relation to the inflammatory reaction.This subject has been extensively reviewed bySpector (1958). An embarrassingly large numberof substances have been described that couldproduce the increase in capillary permeabilitywhich is observed in the inflammatory reaction.The accompanying table lists the naturallyoccurring substances known to increase vascularpermeability to plasma proteins. It is difficult

to prove that any of these substances areinvolved in the inflammatory reaction. Somelight has been cast on the problem by Sevitt(1957) who showed that the reaction to injuryis a double phenomenon. An immediate in-crease in capillary permeability lasting for lessthan 10 minutes (which can be suppressed byantihistaminics) is followed by a delayed in-crease in permeability which begins about onehour after injury and which is unaffected byantihistaminics. Spector and Willoughby (1959)clearly demonstrated the participation of his-tamine in the early stages of inflammation byshowing that previous body depletion of hista-mine by administration of compound 48/80(a powerful histamine liberator) delayed theonset of increased capillary permeability causedin rats by the intrapleural injection of turpentine.There is some evidence from the work ofSpector (1956) and Spector and Willoughby(1960) that activation of globulin-peptidesystems such as the kallikrein-kallidin systemmight be involved in the delayed increase inpermeability.

Little is known about the mode of action ofsubstances which increase vascular permeabilityto plasma proteins. Nearly all these substancesshare in common the property of producingcontraction of smooth muscle cells. It is tempt-

by copyright. on D

ecember 18, 2020 by guest. P

rotectedhttp://pm

j.bmj.com

/P

ostgrad Med J: first published as 10.1136/pgm

j.41.477.425 on 1 July 1965. Dow

nloaded from

July, 1965 ASSCHER and JONES: Capillary Permeability to Plasma Proteins

ing to suggest, therefore, that they might actby producing contraction of endothelial cellswith the consequent appearance of gaps bet-ween the cells or enlargement of alreadyexisting pores. This suggestion was originallymade by Zweifach (1953) who studied theeffect of histamine on the capillaries of themesoappendix of the rat. Majno and Palade(1961) showed that under the influence ofhistamine and 5-hydroxytryptamine, large gapsappeared between the endothelial cells throughwhich mercuric sulphide particles rapidlyescaped. These findings support the view thatthe increase in vascular permeability producedby histamine and 5-hydroxytryptamine mightbe due to contraction of endothelial cells. Thismay not be the sole mechanism whereby thesesubstances act. Thus, Alksne (1959) showedthat histamine produced increased pinocytoticactivity in the endothelial cells. Moreover,histamine produces dilatation of terminalarterioles which raises hydrostatic pressure andincreases capillary -filtration of proteins. Thedilatation of the terminal arterioles may alsoincrease the number of perfused capillaries andso enlarge the surface area available for transferof plasma proteins.The mode of action of angiotensin on

vascular permeability has recently been studiedby Giese (1964). He showed that angiotensininfusion in the rat produced an alternatingpattern of constriction and dilatation in themesenteric arteries similar to that observedby Byrom (1954) in severe experimental hyper-tension. He noted that carbon particles wereinvariably deposited in the dilated segments.The 'stretched pore' phenomenon might, there-fore, be involved in producing the increase inpermeability. Alternatively, the increase inpermeability might be due to anoxia consequenton the arterial constrictions since anoxia isknown to increase vascular permeability toplasma proteins (Landis and others, 1932) andit is also known to produce intracellularfenestrations (Luft and Hechter, 1957). Thereare no data published on the mode of actionof any of the other substances which increase

Capillary Permeability in DiseaseIn view of the difficulties inherent in the

measurement of capillary permeability it is,perhaps, not surprising that data relating tochanges in permeability in disease are relativelyscanty, and, in many instances, probably un-reliable. There is little doubt, however, thatchanges in vascular permeability do occur.They are particularly well documented follow-

ing injury, be this due to direct trauma, thermaland chemical burns, exposure to irritant gases(see Yoffey and Courtice, 1956; Florey, 1962)or irradiation (Wish, Furth, Sheppard andStorey, 1952). Similarly, there is little doubtregarding the significance of increased vascularpermeability in the pathogenesis of allergicoedema (Florey, 1962), or of increased glomeru-lar permeability in the pathogenesis of thenephrotic syndrome (Squire, Hardwicke andSoothill, 1963). Increased permeability to pro-teins has also been postulated in post-decom-pression shock (Brunner, Frick and Biihlmann,1964) and in high-altitude pulmonary oedema(Singh, Kapila, Khanna, Nanda and Rao, 1965),while profound changes in capillary permeabilityhave been demonstrated in occasional cases ofidiopathic oedema (Clarkson, Thompson,Horwith and Lucky, 1960; Weinbren, 1963).

In other conditions, increased capillary per-meability is more doubtful. Some authors havepostulated a generalized increase in capillarypermeability in acute glomerulo-nephritis, butWarren and Stead (1944) found the proteincontent of the oedema fluid to be no higherthan in cardiac failure, and the periorbitaldistribution of oedema may be explained onother grounds (Burch, 1940). Excessive proteinleakage has also been reported by severalauthors in the later stages of normal pregnancy,and in toxemia of pregnancy. (Szontagh, 1949and 1950; Herold, 1950; Burger, 1951; Heroldand Brautigam, 1955; Friedberg and Lutz, 1963).

Schurmann and MacMahon (1933) postulatedthat leakage of plasma proteins into the medialmuscle coat of arteries was responsible forthe appearance of fibrinoid necrosis in hyper-tensive disease. Some support for this view wasobtained by Asscher and Anson (1963a) whoshowed that renal cortical extract inducedboth leakage of plasma proteins and fibrinoidnecrosis when injected into nephrectomized andrenal-hypertensive rats. Since fibrinoid necrosisis commonly considered to be the histologicalhallmark of malignant hypertension, it ispossible that the transition from the benignto the malignant phase is determined by achange in endothelial permeability to plasmaproteins.

These few examples are intended to illustrateboth the variety of disorders which may beassociated with alterations in vascular per-meability, and the paucity of information avail-ale. The methods used in many instances donot allow anything other than tentative con-clusions to be drawn, and it is clear that furtherprogress in this field must await the develop-

431by copyright.

on Decem

ber 18, 2020 by guest. Protected

http://pmj.bm

j.com/

Postgrad M

ed J: first published as 10.1136/pgmj.41.477.425 on 1 July 1965. D

ownloaded from

432 POSTGRADUATE MEDICAL JOURNAL July, 1965

ment of reliable techniques of measuring capil-lary permeability in the human subject.We are indebted to Professor Harold Scarborough

for helpful criticism.

REFERENCES

ALKSNE, J. F. (1959): The Passage of ColloidalParticles across the Dermal Capillary Wall underthe Influence of Histamine, Quart. J. exp. Physiol.,44, 51.

ARESKOG, N. H., ARTURSON, G., GROTrE, G., andWALLENIUS, G. (1964): Studies on Heart LymphII. Capillary Permeability of the Dog's Heart.Using Dextran as a Test Substance, Acta. physiol.scand., 62, 218.

ASSCHER, A. W., and ANSON, S. G. (1963a): AVascular Permeability Factor of Renal Origin,Nature (Lond.), 198, 1097.

ASSCHER, A. W., and ANSON, S. G. (1963b): 'AVascular Permeability Factor of Renal Origin' inProceedings of 2nd International Congress ofNephrology, Prague, 1963. Published by ExcerptaMedica Foundation. pp. 832-834.

BENNETT, N. S., LUFT, J. H., and HAMPTON, J. C.(1959): Morphological Classification of VertebrateBlood Capillaries, Amer. J. Physiol., 196, 381.

BERALDO, W. T. (1955): Substance U and RelatedSubstances from Urine. In: Polypeptides whichStimulate Plain Muscle. ed. by J. H. Gaddum.pp. 58-66. London: E. & S. Livingstone.

BREWER, D. B. (1951): Renal Clearances of Dextransof Varying Molecular Weights, Proc. roy. Soc.Med., 44, 561.

BROOKS, S. A., DAVIES, J. W. L., GRABER, I. G., andRICKETTS, C. R. (1960): Labelling of Inulin withRadioactive Iodine, Nature (Lond.), 188, 675.

BRUNNER, F. P., FRICK, P. G., BUHLMAN, A. A. (1964):Post-decompression Shock due to Extravasation ofPlasma, Lancet, i, 1071.

BURCH, G. E. (1940): Formation of Oedema in theEyelids of Man, A.M.A. Arch. intern. Med., 65,477.

BURGER, H. (1951): Uber die Beeinflussung derKapillarresistenz und der Kapillar Permeabilitat inder Schwangerschaft durch Rutin, Z. Geburtsh.Gynik, 135, 182.

BYROM, F. B. (1954): The Pathogenesis of Hyper-tensive Encephalopathy and its Relation to theMalignant Phase of Hypertension, Lancet, ii, 201.

CAREY, M. J., and CONWAY, E. J. (1954): Comparisonof Various Media for Immersing Frog Sartorii atRoom Temperature and Evidence for RegionalDistribution of Fibre Na+., J. Physiol. (Lond.).125, 232.

CLARKSON, B., THOMPSON, D., HORWITH, M., andLUCKY, E. M. (1960): Cyclical Oedema and ShockDue to Increased Capillary Permeability, Amer. J.Med., 29, 193.

COHEN, S., FREEMAN, T., and MCFARLANE, A. S.(1961): Metabolism of ''I-labelled HumanAlbumin, Clin. Sci., 20, 161.

COPE, 0., and LITWIN, S. B. (1962): Contribution ofthe Lymphatic System to the Replenishment of thePlasma Protein following a Haemorrhage, Ann. Surg.156, 655.

COPE, 0., and MOORE, F. D. (1944): Study of Capil-lary Permeability in Experimental Burns and BurShock using Radioactive Dyes in Blood and Lymph,J. clin. Invest., 23, 241.

COURTICE, F. C. (1961): The Transfer of Proteinsand Lipids from Plasma to Lymph in the Leg ofthe Normal and Hypercholesterolamic Rabbit, J.Physiol. (Lond.), 155, 456.

COURTICE, F. C., and MORRIS, B. (1955): The Ex-change of Lipids between Plasma and Lymph ofAnimals, Quart. J. exp. Physiol., 40, 138.

CREESE, R., D'SILVA, J. L., and SHAW, D. M. (1962).Interfibre Fluid from Guinea-Pig Muscle, J. Physiol.(Lond.), 162, 44.

CREESE, R., and NORTHOVER, J. (1951): Maintenanceof Isolated Diaphragm with Normal Sodium Con-tent, J. Physiol., 155, 343.

CROXATrO, H. (1955): Active Polypeptides Formedby Pepsin. In: Polypeptides which Stimulate PlainMuscle. ed. by J. H. Gaddum. pp. 92-102. London:E. & S. Livingstone.

CULLUMBINE, H., and RYDON, H. N. (1946): A Studyof the Formation, Properties and Partial Purifica-tion of Leucotaxine, Brit. J. exp. Path., 27, 33.

DRINKER, C. K., and FIELD, M. E. (1931): TheProtein Content of Mammalian Lymph and theRelation of Lymph to Tissue Fluid, Amer. J.Physiol., 97, 32.

DRINKER, C. K., FIELD, M. E., HEIM, J. W., andLEIGH, O. C. (1934): The Composition of OedemaFluid and Lymph in Oedema and ElephantiasisResulting from Lymphatic Obstruction, Amer. J.Physiol., 109, 572.

DRINKER, C. K., and YOFFEY, J. M. (1941): Lym-phatics, Lymph and Lymphoid Tissue, Harvard:University Press.

DUTHIE, E. S., and CHAIN, E. (1939): A PolypeptideResponsible for Some of the Phenomena of AcuteInflammation, Brit. J. exp. Path., 20, 412.

ELDER, J. H., and WILHELM, D. L. (1958): Enzyme-like Globulins from Serum Reproducing theVascular Phenomenon of Inflammation V. Activ-able Permeability Factor in Human Serum, J. exp.Path., 39, 335.

EULER, U. S V., and GADDUM, J. H. (1931): AnUnidentified Depressor Substance in Certain TissueExtracts, J. Physiol. (Lond.), 72, 74.

FARQUHAR, M. G., WISSIG, S. L., and PALADE, G. E.(1961): Glomerular Permeability. 1. FerritinAcross the Normal Glomerular Capillary Wall, J.exp. Med., 113, 47.

FLODIN, P. (1962): Dextran Gels and their Applica-tion in Gel Filtration. Halmstad: MeijelsBokindustri.FLOREY, H. W. (1962): General Pathology, 3rd Ed.

pp. 30-37. London: Lloyd-Luke.FLOREY, H. W. (1964): The Transport of Materials

Across the Capillary Wall, Quart. J. exp. Physiol.,49, 117.FREY, E. K., K trrT, H., and WERLE, E. (1950):Kallikrein (Padutin). Ed. Enke, Stuttgart: Bahl.FRIEDBERG, V and LTrz, J. (1963): Untersuchungeniber die Capillarpermeabilitat in der Schwanger-schaft, Arch. Gynik., 199, 96.GADDUM, J. H. (1948): Histamine, Brit. med. J., i,867.GIESE, J. (1963): Pathogenesis of Vascular DiseaseCaused by Acute Renal Ischaemia, Acta. path.microbiol. scand., 59, 417.

GIESE, J. (1964): Acute Hypertensive Vascular Dis-ease. 2. Studies on Vascular Reaction Patternsand Permeability Changes by Means of VitalMicroscopy, Acta. path. microbiol scand., 62, 497.

GROTrE, G. (1956): Passage of Dextran Moleculesacross the Blood-lymph Barrier, Acta. chir. scand.Suppl., 211, 1.

by copyright. on D

ecember 18, 2020 by guest. P

rotectedhttp://pm

j.bmj.com

/P

ostgrad Med J: first published as 10.1136/pgm

j.41.477.425 on 1 July 1965. Dow

nloaded from

July, 1965 ASSCHER and JONES: Capillary Permeability to Plasma Proteins 433

GROTTE, G., JUHLIN, L., and SANDBERG, N. (1960):Passage of Solid Spherical Particles across theBlood-lymph Barrier, Acta. physiol. scand., 50, 287.

HARDWICKE, J. (1965): Clin. chim. Acta. In Press.HARDWICKE, J., HULME, B., JONES, J. H., and

RICKETTS, C. R. (1964): The Estimation of RenalPermeability to Radioactively Labelled Polydis-perse Macromolecules on Sephadex Chromato-graphy. Communication to Renal Association.

HARDWICKE, J., and SQUIRE, J. R. (1955): The Re-lationship between Plasma Albumin Concentrationand Protein Excretion in Patients with Proteinuria,Clin. Sci, 14, 509.

HECHTER, 0., and LESTER, G. (1960): Cell Permeabil-ity and Hormone Action, Recent Progr. HormoneRes., 16, 139.

HEROLD, L. (1950): Das Permeabilititsproblem inder Schwangenschaft, Z. Geburtsh. Gynik., 132,145.

HEROLD, L. and BRAUTIGAM, H. H. (1955): Klinische-experimentelle Utersuchungen iiber das Verhaltender Capillarpermeabilitat by Spatgestosen, Arch.Gyniik., 187, 83.

HILTON, S. M., and LEWIS, G. P. (1955): TheMechanism of the Functional Hyperamia in theSubmandibular Salivary Gland, J. Physiol (Lond.),129, 253.

HOLDSTOCK, D. J., MATHAS, A. P., and SCHACHTER, M.(1957): A Comparative Study of Kinin, Kallidinand Bradykinin, Brit. J. Pharmacol., 12, 149.

JONES, J. H., and PETERS, D. K. (1965): Unpublisheddata.

KLATZO, I., MIQUEL, J., and OTENASEK, R. (1963):The Application of Fluorescein Labelled SerumProteins to the Study of Vascular Permeability inthe Brain, Acta neuropath. (Berl.), 2, 144.

LANDIS, E. M. (1927): Micro-injection Studies ofCapillary Permeability II. Relation between Capil-lary Pressure and Rate at which Fluid Passesthrough the Walls of Single Capillaries, Amer. J.Physiol., 82, 217.

LANDIS, E. M. (1946): Capillary Permeability and theFactors Affecting the Composition of CapillaryFiltrate, Ann. N.Y. Acad. Sci., 46, 713.

LANDIS, E. M., JONAS, L., ANGEVINE, N., and ERB, W.(1932): The Passage of Fluid and Protein throughthe Human Capillary Wall during Venous Conges-tion, J. clin. Invest. 11, 171.

LANDIS, E. M., and PAPPENHEIMER, J. R. (1963):Exchange of Substances through the CapillaryWalls. In: Handbook of Physiology. Section 2.Circulation. pp. 961-1034. Washington D.C.: Amer.Physiol. Soc.

LEWIS, T. (1927): The Blood Vessels of the HumanSkin and their Responses. ed. Shaw, London.

LUFr, J., and HECHTER, 0. (1957): An ElectronMicroscopic Correlation of Structure with Func-tion in the Isolated Perfused Cow Adrenal. Pre-liminary Observations, J. Biophys. Biochem. Cytol.,3, 615-621.

MAJNO, G., and PALADE, G. E. (1961): Studies onInflammation. 1. The Effects of Histamine andSerotonin on Vascular Permeability. An ElectronMicroscopic Study, J. Biophys. biochem. Cytol, 11,571.

MASSON, G. M. C.. PLAHL. G., CORCORAN, A. C.,PAGE, I. H. (1953): Accelerated Vascular Diseasefrom Saline and Renin in Nephrectomized Dogs,Arch. Path., 55, 85.

MATTHEWS, C. M. E. (1957): The Theory of TracerExperiments with '3I-labelled Plasma Proteins,Phys. in Med. Biol., 2, 36.

MAYERSON, H. S. (1963): The Physiological Import-ance of Lymph. In: Handbook of Physiology. Sec-tion 2. Circulation. pp. 1035-1073. WashingtonD. C.: Amer. Physiol. Soc.

MAYERSON, H. S., WOLFRAM, C. G., SHIRLEY, H. H.,and WASSERMAN, K. (1960): Regional Differencesin Capillary Permeability, Amer. J. Physiol., 198,155.

MENKIN, V. (1940): Dynamics of Inflammation. NewYork: Macmillan.

MENKIN, V. (1956): Biochemical Mechanisms in In-flammation. 2nd ed. Springfield, Illinois: CharlesC. Thomas.

MILES, A. A., and MILES, E. M. (1952): VascularReactions to Histamine, Histamine-Liberator andLeucotaxine in the Skin of Guinea-pigs, J. Physiol.(Lond.), 118, 228.

MILES, A. A., and WILHELM, D. L. (1955): Enzyme-like Globulins from Serum Reproducing the Vas-cular Phenomena of Inflammation. 1. An ActivablePermeability Factor and its Inhibitor in Guinea-Pig Serum, Brit. J. exp. Path., 39, 343.

MILL, P. J. (1964): Induction of a Vascular Pet-meability Factor by the Interaction of Guinea PigTissue Debris with Serum, Nature (Lond.), 185, 246.

MILL, P. J., ELDER, J. M., MILES, A. A., andWILHELM, D. L. (1958): Enzyme-like Globulinsfrom Serum Reproducing the Vascular Phenomenonof Inflammation. VI. Isolation and Properties ofPermeability Factor and its Inhibitors in HumanPlasma, Brit. J. exp. Path., 39, 343.

MOORE, D. M., and RUSKA, H. (1957): The FineStructure of Capillaries and Small Arteries, J.biophys. biochem. Cytol., 3, 457.

PALADE, G. E. (1953): Fine Structure of BloodCapillaries, J. Appl. Physics, 24, 14.

PALADE, G. E. (1956): The Endoplasmic Reticulum,J. biophys. biochem. Cytol., 2, No. 4 Suppl. 85.PALADE, G. E. (1960): Transport in Quanta across

the Endothelium of Blood Capillaries, Anat. Rec.,136, 254.PAPPENHEIMER, J. R. (1953): Passage of Molecules

through Capillary Walls, Physiol. Rev., 33, 387.PAPPENHEIMER, J. R., and SOTO-RIVERA, A. (1948):Effective Osmotic Pressure of the Plasma Proteinsand other Quantities Associated with the CapillaryCirculation in the Hindlimbs of Cats and Dogs,Amer. J. Physiol., 152, 471.

PERLMANN, G. E., GLENN, W. W. L., and KAUFMAND. (1943): Changes in the Electrophoretic Patternin Lymph and Serum in Experimental Bums, J.diin. Invest., 22, 627.

PERNOW, B. (1955): Distribution and Properties ofSubstance P. In: Polypeptides which StimulatePlain Muscle, ed. by J. H. Gaddum. pp. 28-38.London: E. & S. Livingstone.

PORATH, J., and FLODIN, P. (1959): Gel Filtration:A Method for Desalting and Group Separation,Nature (Lond.), 183, 1657.

RHODIN, J. A. G. (1962): The Diaphragm of CapillaryEndothelial Fenestrations, J. Ultrastr. Res., 6, 171.

ROCHA e SILVA, M. (1955): Bradykinin: Occurrenceand Properties. In: Polypeptides which StimulatePlain Muscle, ed. by J. H. Gaddum. pp. 45-47.London: E. & S. Livingstone.

ROCHA e SILVA, M., and DRAGSTEDT, C. A. (1941):Observations on the Trypan Blue Capillary Per-meability Test in Rabbits, J. Pharmacol. exp. Ther.,73, 405.

by copyright. on D

ecember 18, 2020 by guest. P

rotectedhttp://pm

j.bmj.com

/P

ostgrad Med J: first published as 10.1136/pgm

j.41.477.425 on 1 July 1965. Dow

nloaded from

434 POSTGRADUATE MEDICAL JOURNAL July, 1965

ROWLEY, D. A., and BENDITT, E. P. (1956): 5-Hydroxytryptamine and Histamine as Mediators ofthe Vascular Injury Produced by Agents whichDamaged Mast Cells in Rats, J. exp. Med., 103,399.

SCHURMANN, P., and MACMAHON, H. E. (1933): DieMaligne Nephrosklerose, Zugleich ein Beitrag zurFrage der Bedeutung der Bentgewebsschrante,Virchows. Arch. path. Anat., 291, 47.

SEVITT, S. (1957): Early and Delayed Oedema andIncrease in Capillary Permeability after Burs ofthe Skin, J. Path. Bact., 75, 27.

SHIRLEY, H. H., WOLFRAM, C. G., WASSERMAN, K., andMAYERSON, H. S. (1957): Capillary Permeability toMacromolecules: Stretched Pore Phenomenon,Amer. J. Physiol., 190(2), 189.

SINGH, I., KAPILA, C. C., KHANNA, P. K., NANDA,R. B., and RAO, B. D. P. (1965): High AltitudePulmonary Oedema, Lancet, i, 229.

SOOTHILL, J. F. (1962): The Estimation of EightSerum Proteins by a Gel Diffusion PrecipitinTechnique, J. Lab. clin. Med., 59, 859.

SPECTOR, W. G. (1951): The Role of Some HigherPeptides in Inflammation, J. Path. Bact., 63, 93.

SPECTOR, W. G. (1956): The Mediation of AlteredCapillary Permeability in Acute Inflammation, J.Path. Bact., 72, 367.

SPECTOR, W. G. (1958): Substances which AffectCapillary Permeability, Pharmacol. Rev., 10, 475.

SPECTOR, W. G., and WILLOUGHBY, D. A. (1957):Capillary Permeability Factors, Nucleosides andHistamine Release, J. Path. Bact., 73, 135.

SPECTOR, W. G., and WILLOUGHBY, D. A. (1959): TheDemonstration of the Role of Mediators inTurpentine Pleurisy in Rats by Experimental Sup-pression of the Inflammatory Changes, J. Path.Bact., 77, 1.

SPECTOR, W. G., and WILLOUGHBY, D. A. (1960):The Suppression by Anti-esterases of IncreasedCapillary Permeability in Acute Inflammation, J.Path. Bact., 79, 21.

SQUIRE, J. R., HARDWICK, J., and SOOTHILL, J. F.(1962): "Proteinuria". In: Renal Disease ed. byD.A.K. Black. pp. 213-230. Oxford: BlackwellScientific Publications.

STARLING, E. H. (1896): On the Absorption ofFluid from the Connective Tissue Spaces, J. Physiol.(Lond), 19, 312.

STEWART, P. B., and BLISS, J. Q. (1957): The Per-meability Increasing Factor in Diluted HumanPlasma, J. Exper. Path., 38, 462.

SZONTAGH, F. E. (1949): Capillary Permeability, SerumProteins and Hamatocrit Values in Normal Preg-nancy, Gynaecologia (Basel), 127, 240.

SZONTAGH, F. E. (1950): Capillary Premeability inLate Pregnancy Toxaemias, Gyncecologia (Basel),130, 109.

WALLENIUS, G. (1954): Renal Clearance of Dextranas a Measure of Glomerular Permeability, Acta.Soc. Med. upsalien., 59, Supp. 4.

WALTER, M., N-I., and WILLOUGHBY, D. A. (1965):The Effect of Tissue Extracts on Vascular Per-meability and Leucocyte Emigration, J. Path Bact.89, 255.

WARREN, J. V., and STEAD, E. A. (1944): ProteinConcentration of Tissue Fluids in Acute Nephritis,Amer. J. med. Sci., 208, 618.

WASSERMAN, K., JOSEPH, J. D., and MAYERSON, H. S.(1956): Kinetics of Vascular and ExtravascularProtein Exchange in Unbled and Bled Dogs, Amer.J. Physiol., 184, 175.

WASSERMAN, K., LOEB, L., and MAYERSON, H. S.(1955): Capillary Permeability to Macromolecules,Circulat. Res., 3, 594.

WASSERMAN, K., and MAYERSON, H. S. (1951): Ex-change of Albumin between Plasma and Lymph,Amer. J. Physiol., 165, 15.

WEINBREN, I. (1963): Spontaneous Periodic Oedema,Lancet, ii, 544.

WERLE, E. (1955): The Chemistry and Pharmacologyof Kallikrein and Kallidin. In: Polypeptides whichStimulate Plain Muscle. ed. by J. H. Gaddum.pp. 20-27. London: E. & S. Livingstone.

WILHELM, D. L., MILES, A. A., and MACKAY, M. E.(1955): Enzyme-like Globulins from Serum Re-producing the Vascular Phenomena of Inflamma-tion. II. Isolation and Properties of thePermeability Factor and its Inhibitor, Brit. J. exp.Path., 36, 82.

WINTERNITZ, M. C., MYLON, E., WATERS, L. L., andKATZENSTEIN, R. (1940): Studies on the Relationof the Kidney to Cardiovascular Disease, Yale I.Biol. Med., 12, 623.

WISH, L., FURTH, J., SHEPPARD, C. W., and STOREY,R. H. (1952): Disappearance Rates of TaggedSubstances from the Circulation of Roentgen Ir-radiated Animals, Amer. J. Roentgenol., 67, 628.

WISSIG, S. L. (1958): An Electron Microscopy Studyof the Permeability of Capillaries in Muscle, Anat.Rec., 130, 467.

YOFFEY, J. M., and COURTICE, F. C. (1956): Lym-phatics, Lymph and Lymphoid Tissue. 2nd ed.London: Edward Arnold.

ZWEIFACH, B. W. (1953): Analysis of the Inflam-matory Reaction through the Response of theTerminal Vascular Bed to Micro-trauma. In:The Mechanism of Inflammation, ed. by G. Jasminand A. Robert, pp. 77-84, Montreal: Acta Inc.

by copyright. on D

ecember 18, 2020 by guest. P

rotectedhttp://pm

j.bmj.com

/P

ostgrad Med J: first published as 10.1136/pgm

j.41.477.425 on 1 July 1965. Dow

nloaded from